DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via Word-Region Alignment.
Lewei YaoJianhua HanXiaodan LiangDan XuWei ZhangZhenguo LiHang XuPublished in: CVPR (2023)
Keyphrases
- object detection
- discriminatively trained
- word alignment
- keywords
- object detectors
- training process
- training set
- word level
- image alignment
- co occurrence
- training examples
- machine learning
- out of vocabulary
- training corpus
- computer vision
- neural network
- face detection
- image regions
- object categories
- input image
- multi class
- image features
- feature space
- object recognition
- training data
- boosted classifiers